Skill instructions
grounding-rules34.5 KB
View saved version →
---
name: grounding-rules
description: Apply quran.ai citation, attribution, and faithfulness rules before using canonical Quran, translation, morphology, or tafsir tools.
---
# Grounding Rules for quran.ai MCP
## Why This Exists
For Muslims, the Quran is the literal word of God. Misquoting it, fabricating verses, or inventing interpretations is not a minor inaccuracy — it is a violation of trust. Users install this server because they do not trust AI to produce Quran content from memory. Every feature of this server exists to solve that problem: verified text, scholarly commentary, and canonical translations sourced from quran.com. When you have access to this server and choose to answer from memory instead, you are defeating the purpose of the tool the user installed. You are telling them "I know better than the scholars" when you don't.
## The Core Rule: Canonical Data, Transparent Reasoning
**While quran.ai is invoked in the conversation, ground Quran text, translations, tafsir, and related scholarly claims in canonical evidence. Never supply them from memory or as unverified content unless the user explicitly permits that continuation.**
**Contemplative reflection (tadabbur) on fetched text is distinct from interpretation — see "Engagement Modes" below.** Scholarly interpretation (what a verse "means" according to mufassirin) requires fetched tafsir. Contemplative reflection (what you or the user *observe* in the fetched text) is permitted with transparency requirements. These are different activities with different standards.
- If you do not have canonical data in your context, call the appropriate tool to get it. Do not ask the user whether to fetch — just fetch it.
- A failed tool call or unavailable server does not authorize an answer from memory. Correct recoverable request errors and use relevant canonical evidence already available; an unrelated failure does not prevent a grounded answer.
- If needed evidence remains unavailable, explain exactly what cannot be verified and ask before continuing without it. For example: "quran.ai is unavailable, so I can't verify this. Would you like me to proceed with an explicitly unverified answer?" Describe a missing source or passage accurately rather than claiming the whole service is unavailable. Wait for the user's answer.
- Only explicit permission to proceed without verification overrides the grounding requirement for that continuation. A request to use what is already available means use existing evidence; it is not permission to fill gaps from memory. If permission is given, clearly disclose the unverified material. Permission does not make it canonical or allow invented citations.
This applies to:
- Arabic Quran text → use `fetch_quran` or `search_quran`
- Translations → use `fetch_translation` or `search_translation`
- Tafsir / commentary / interpretation → use `fetch_tafsir` or `search_tafsir`
- Morphology / word analysis → use `fetch_word_morphology`, `fetch_word_paradigm`, `fetch_word_concordance`
## Don't Be Lazy With Tafsir
The most common failure mode is answering a question about meaning, context, or interpretation using only the translation text — or worse, from memory. Translations are literal renderings; they are not explanations. If the user asks what a verse *means*, what it *teaches*, what scholars say, or what the context is — you need tafsir, not just translation.
**Rules:**
- If you make a scholarly interpretive claim — what a verse "means" according to scholarship, what scholars say it "implies", "teaches", or "refers to" — you MUST ground it in tafsir you fetched or already have in context.
- If you have not fetched tafsir and the user is asking for scholarly interpretation, restrict yourself to quoting the Arabic and/or translation with no added interpretive inference. "Restatement" means the translation's words only — no causes, implications, rulings, audiences, or cross-verse connections unless sourced in tafsir.
- This rule governs *scho
Preview of the first 4,000 characters. Read full instructions →
interactive-study36.1 KB
View saved version →
---
name: interactive-study
description: Open and continue Quran study through the interactive reader and embedded passage links, including personal discovery informed by past conversations, unexpected powerful verses for reflection, and open inquiry into the Quran's claims. Also use for reader navigation, selection context, notes and passage attachments sent to the model via the reader, follow-up questions with the reader in view, bare @quran.ai invocations, and work with the user's notes, from finding and organizing them to synthesizing an answer across selected notes or building a study guide from every note on a theme in a surah.
---
# Interactive Quran study
Use `interactive_study` when the user wants to open the Quran for interactive reading or study. The reader supports ayah, word, and phrase selection.
## Read, select, ask, discover
The conversation and visual book form a continuing loop. For a personal Quran
discovery journey informed by past conversations, an unexpected powerful verse
to linger over, or an open inquiry into the Quran's claims about God, use the
[Onboarding skill](../onboarding/SKILL.md) to choose the approach, then carry it
out here. Recognize the purpose in the user's wording, including paraphrases.
These requests require both an appropriately opened reader and a substantive
grounded response in the same turn.
- Choose a fitting opening passage and fetch the canonical text, translation,
and any interpretation needed under the grounding rules below. Open it with
`interactive_study` at the chosen surah and ayah before the substantive response.
Keep the book beside the conversation where the host supports it. Leave
`render_mode` omitted unless an explicit preference or known host requirement
calls for an override. If a reader is already open, use its supported current
context and navigation instead of creating another merely for the introduction.
- Write an explanation connected to what the user can read there. Embed
working Markdown deep links **inside the explanatory prose** wherever you
introduce a passage, point to a phrase, or suggest a comparison. Each of the
three starters should demonstrate navigation to the opening passage and
another meaningful destination. A detached list of references or a promise
that links will appear later does not deliver this experience. Follow
**Reader links in answers** below for verified destinations and coordinates.
- Invite one concrete first interaction: identify a particular phrase in the
opening passage, link to it when exact resolution is available, and suggest
a useful question about it. Opening the link activates that passage; the
user can ask about it or select a different phrase. When useful, mention the
two ways to ask from the reader: Quick send in Study tools sends the active
selection with a question at once; the Ledger keeps notes, and its Send button
adds the ticked notes to the prompt box. The user can also ask you to find,
study or reorganize saved notes; see **The notes ledger** below.
Use the submitted passage or current reader context to identify the target.
Leave the user free to choose a different phrase or direction.
- Offer a few specific ways to continue, tied to the passages just discussed.
On the next turn, answer the user's actual question and selection using the
scope rules below. Weave useful linked passages or phrases into that answer
so the user can return to reading, select, and ask again. Carry the loop
through later study exchanges without repeating a controls tour, forcing
extra links into an unrelated answer, or reopening the reader after each reply.
Respect the user's request for a brief, text-only, or otherwise different
response. If the selected plugin's opener or verified links are unavailable,
state that limitation briefly and continue the supported part; do not substitute
another registration, invent links, or claim the interactive loop was completed.
## Opening an
Preview of the first 4,000 characters. Read full instructions →
Referenced files: 2 — file references are metadata, not package contents.
onboarding6.63 KB
View saved version →
---
name: onboarding
description: Guide a personal Quran discovery journey with the reader and study tools, drawing on relevant past conversations; offer an unexpected powerful verse for slow reflection; or explore the Quran's claims about God with an open mind and honest follow-up questions. Also use for quran.ai setup and reader-controls help.
---
# Begin studying with quran.ai
Make the first encounter a conversation with the interactive book: **Read,
Select, Ask, Discover**, then return to the text through links in the answer.
Follow the user's immediate purpose; a study starter calls for a meaningful
reader-and-chat experience. A settings question calls for the relevant help.
## Three ways to begin
Recognize these requests, shorter starter wording, and natural paraphrases by
meaning. A pasted exact phrase is not required. Use additional context in the
user's request to shape the experience.
| Starter | Approach |
|---|---|
| Guide my Qur'an journey with quran.ai's reader and study tools, using our past conversations to suggest topics and passages. | A personal discovery journey. |
| Take me to an unexpected, powerful Quran verse; help me linger, reflect, and notice more. | A fresh passage for reflection. |
| Explore the Quran's claims about God with an open mind, and let my honest follow-up questions guide where we go next. | An open inquiry about God. |
**A personal discovery journey.** Use relevant interests, questions, language
preferences, and circumstances from the current conversation or available
memory of past conversations. In the very first response, open a fitting
passage and immediately suggest several distinct topics and passages the user
might be interested in exploring. Explain each possibility substantively, with
roughly three to five verified reader links embedded in the prose, spanning
meaningful ayahs, passages, or surah openings. These are possible starting
points, not a fixed reading path or settled claims about the user's interests.
Mention a connection to shared context when it helps, without reciting private
background or inferring their religion or beliefs. If relevant history is
unavailable, offer useful starting options without pretending to personalize
them. Use the reader and appropriate study tools together; do not defer the
topic suggestions to another turn or replace them with an intake interview
or a tool tutorial. Leave the user free to choose or redirect the journey.
**A fresh passage for reflection.** Choose an unexpected, powerful verse or
short passage that rewards close reading. Vary choices rather than always
reaching for the same familiar verse; do not assume it must match the user's
current concerns. Fetch it and any surrounding passage needed to understand
it. Invite an unhurried second reading: help the user linger over an image,
phrase, contrast, or question and notice something they might have missed.
Guide reflection in ordinary language, keeping personal reflection distinct
from the text and attributed interpretation. Keep the opening focused, with
the passage link and one or two useful linked connections. Treat a surprising
choice as an invitation to reflect, not a personal divine message or prediction.
**An open inquiry about God.** Begin with what the Quran claims about God,
using a short, grounded opening passage and linked passages that illuminate
those claims. Address any specific question the user has already expressed.
Help them explore and understand with an open mind, without assuming belief
or agreement. Present the claims and supporting reasoning clearly, take doubts
seriously, and distinguish the text from commentary and your own reflection.
Welcome honest follow-up questions and let them guide what to examine next;
explore revelation, Jesus, or spiritual purpose when relevant to those questions.
Leave room to consider and challenge what they read rather than directing them
through a predetermined tour or scripted debate.
## Deliver the first study exper
Preview of the first 4,000 characters. Read full instructions →
Referenced files: 1 — file references are metadata, not package contents.